Competition Stories: July to October 2021

Welcome to the Competition Stories – a bimonthly exploration of recent courts and competition law agencies’ decisions. Authored by Makis Komninos, a renowned expert in the field, this new column aims to go through the latest and most important developments in competition law of the last two months. We call them “stories” because Makis has promised to include some anecdotes from time to time, and not just stay at the black letter. Enjoy!


What a period this has been for EU competition law developments! There were no “Stories” for July-August, although there were a few minor developments, which are mentioned below. However, September and October did not fail to excite.

First, it became public that the judgment in Google Shopping will be available on November 10 (stay tuned for the end of year “Stories”!). Second, I cannot ignore the five-day Google Android hearing at the General Court on September 27 – October 1. It brought back memories to me from 15 years ago, when the then CFI had again dedicated five days to the Microsoft hearing.

It’s interesting to see how long these mega-litigations take from start to finish:

In the Shopping case, the application for annulment was filed on September 11, 2017. So the litigation will have taken 4 years and 2 months. In the Android case, the application for annulment was filed on October 9, 2018. So it took almost three years for the hearing. When will the judgment come out? Let me speculate a bit. The case was heard by the Sixth Chamber (extended composition), which includes at least one judge whose mandate expires on August 31, 2022. Bearing in mind that the Member State concerned has already published an invitation for the appointment of a new judge, this means that the current member’s mandate will not be renewed. Therefore, in all likelihood, the judgment will come out before that date (there are theoretical scenarios that the judgment could take longer and be rendered by a 3-member Chamber but I think we can discard them in this case). Since I find it very unlikely for the judgment to be out in August (!), my speculative guess is that we should expect it sometime in April-June 2022. So the litigation will probably have lasted less than 4 years, maybe close to 3 years and a half!

And what about Microsoft? The application for annulment was filed on June 7, 2004 and the judgment was rendered on September 17, 2007, so it took 3 years and 3 months. And it was a Grand Chamber judgment!

Enough with statistics. Let’s now go to the latest products from Luxembourg and Brussels.

I. Court of Justice

  1. Sumal

This was a Grand Chamber judgment answering preliminary questions of a Spanish court in a follow-on damages litigation. Spain is at the forefront of such follow-on claims as a result of the Commission’s trucks cartel decision of 2016. The action for damages was brought by Sumal against Mercedes Benz Trucks España (a subsidiary of Daimler AG), to whom the European Commission’s infringement decision was not addressed. The question at stake was whether the doctrine of the single economic unit developed by the Court of Justice provides grounds for extending civil liability in damages from the parent company to the subsidiary (downwards), or it applies solely in the reverse, i.e. in order to extend liability from subsidiaries to the parent company (upwards). AG Pitruzzella’s Opinion supported a “downward” extension of liability and the Court follows the Advocate General, but maybe goes further in some respects.

The judgment has been described as plaintiff-friendly, but this may be a rather superficial reading. Indeed, there may be areas where it is defendant-friendly, as I explain below.

We know from Skanska that although the doctrine of single economic unit and the corresponding attribution of liability is something that was conceived and developed in the context of public enforcement, it also applies to private enforcement and regulates attribution of civil liability in damages. The Court confirms this anew (para. 37 et seq.). Then, after recalling the case law on parental liability, i.e. the “upward” extension of liability from subsidiary to parent (paras 42-44), the Court delves into the question at stake: should there be a “downward” extension of liability from parent to subsidiary and under what conditions, if at all? The Court effectively says “yes” but adopts a middle ground approach, by requiring that certain conditions are present.

According to the Court, a subsidiary can be held liable for the anti-competitive conduct of its parent company if it can be proven by the claimant that these two companies constituted a single economic unit at the time of the infringement (para. 48) by having regard to (i) the economic, organisational and legal links between the parent company and the subsidiary (para. 51) and (ii) “the existence of a specific link between the economic activity of that subsidiary and the subject matter of the infringement for which the parent company was held to be responsible” (para. 51). It is the latter condition, the new component of a “specific link”, where the Court breaks new ground and offers defendants some respite. In offering guidance as to the “specific link”, the Court stresses that “in circumstances such as those at issue in the main proceedings, the victim should in principle establish that the anticompetitive agreement concluded by the parent company, for which it has been punished, concerns the same products as those marketed by the subsidiary” (para. 52). In other words, Sumal would have to prove that the anti-competitive agreement for which Daimler AG was held responsible in the Commission’s decision concerns the same products as those sold by Mercedes Benz Trucks España.

Before reaching that conclusion, the Court makes certain interesting pronouncements in paras 45-47, again breaking new ground. Thus, according to the Court, “the organisation of groups of companies that may constitute an economic unit may be very different from one group to another. There are, in particular, some groups of companies that are ‘conglomerates’, which are active in several economic fields having no connection between them” (para. 45). Therefore, an automatic extension of liability in such cases is not possible. Indeed, “the same parent company may be part of several economic units made up, depending on the economic activity in question, of itself and of different combinations of its subsidiaries all belonging to the same group of companies. If that were not the case, a subsidiary within such a group could be held liable for infringements committed in the context of economic activities entirely unconnected to its own activity and in which they were in no way involved, even indirectly” (para. 47). This part of the judgment can certainly be seen as defendant-friendly. It also raises interesting questions about possible spill-over effects to public enforcement, if, as the Court says, the concept of “undertaking” cannot be different between public and private enforcement. It remains to be seen how the Commission will apply this principle in its own decisions in the future.

The Court includes some helpful guidance on how the subsidiary’s due process rights are to be respected in a situation of downward extension of liability (para. 53 et seq.). These are worth reading but I will stop here. Interestingly, the Court reminds that the subsidiary cannot under any circumstances dispute the finding of an infringement that has been imputed on its parent, since this would not be possible under Article 16(1) of Regulation 1/2003, which provides that national courts cannot take decisions running counter to a Commission decision (para. 55).

  1. Volvo

Again another preliminary reference from Spain in a private enforcement case following on the trucks cartel decision of the Commission. Still, not as foundational as Sumal. RH is an undertaking domiciled in Cordoba (Spain) that purchased five trucks from a Volvo Group España dealer between 2004 and 2009. The action for damages was filed against Volvo entities domiciled in Sweden and Germany, and also against the Spanish subsidiary Volvo Group España, which has its registered office in Madrid. Although the claimant had purchased the vehicles in Cordoba and was domiciled in that city, it brought its claim before the Madrid Commercial Court. The specific question related to Article 7(2) of Regulation 1215/2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters, which provides for a special basis of jurisdiction in matters of tort, delict or quasi-delict, for the courts where the harmful event occurred. The Spanish court was asking, however, not about international jurisdiction per se but rather whether Article 7(2) is a rule purely related to international jurisdiction, or whether it also determines local territorial jurisdiction.

This is a question that has not arisen before in the EU case law, to the best of my knowledge, and I have to say I am a bit puzzled, because this is normally a question that can be resolved under the principle of national procedural autonomy. In any event, the Court of Justice did answer the question and held that it was “it is clear from the very wording of Article 7(2) of Regulation No 1215/2012 that that provision confers directly and immediately both international and territorial jurisdiction on the courts for the place where the damage occurred” (para. 33). That being said, the Court does also say that although “the Member States may not apply criteria for the conferral of jurisdiction which differ from those deriving from Article 7(2)”, still “the delimitation of the court’s jurisdiction within which the place where the damage occurred, within the meaning of that provision, is situated, is as a rule a matter for the organisational competence of the Member State to which that court belongs” (para. 34).

Indeed, so that there is no doubt, the Court clarifies that Member States can decide to centralise jurisdiction before a single specialised court, in the interests of the sound administration of justice. Such rules are quite usual in the area of private antitrust enforcement (e.g. in France). And “the technical complexity of the rules applicable to actions for damages for infringements of competition law provisions may also militate in favour of a centralisation of jurisdiction” (para. 37).

So what’s the answer in this case? In which court should the victim of the cartel bring the action? Where is “the place where the damage occurred”? The Court of Justice relies on its earlier precedent in VKI and stresses that the critical place is where the goods or services at stake were purchased. This should be the case here. However, in the Court’s words, “that approach implies that the purchaser that has been harmed exclusively purchased goods affected by the collusive arrangements in question within the jurisdiction of a single court. Otherwise, it would not be possible to identify a single place of occurrence of damage with regard to the purchaser harmed” (para. 40). In such a case, “the place where the damage occurred is identifiable only for each alleged victim taken individually and is located, in general, at that victim’s registered office” (para. 41). The same goes for the scenario where the products were purchased in places within the jurisdiction of several courts. According to the ECJ, “to confer jurisdiction on the courts for the place where the undertaking harmed has its registered office is consistent with the requirement of predictability […] since the defendants, members of the cartel, cannot be unaware of the fact that the purchasers of the goods in question are established within the market affected by the collusive practices. In addition, that conferral of jurisdiction is consistent with the objective of proximity, and the place of the registered office of the undertaking harmed fully guarantees the efficacious conduct of potential proceedings” (para. 42).

  1. AG Rantos’s Opinion in Volvo and DAF

The trucks cartel production of interesting preliminary references from Spain continues. This is the first Opinion of AG Rantos (Greek Advocate General that originally replaced AG Sharpston after Brexit) in the area of competition law and deals with some important questions relating to the Damages Directive, as the latter was transposed into Spanish law.

The plaintiff in this case had purchased, during 2006 and 2007, three trucks manufactured by Volvo and DAF and brought a follow-on action for damages. His claim was upheld in part by the first instance Spanish court, and Volvo and DAF were ordered to pay compensation amounting to 15% of the purchase price of the trucks. The court rejected the plea raised by the defendants that the action was time-barred, applying the five-year limitation period provided in the new Spanish legislation, which transposed the Damages Directive. In addition, in accordance with that legislation, the national court applied the presumption of harm caused by the infringement at issue and used its power to assess the harm, as provided for in the Damages Directive. The two defendants brought an appeal against that judgment and submitted, first, that the action was time-barred because the one-year limitation period of the Spanish Civil Code, which in their view is applicable, began to run from the publication of the Commission’s press release on July 19, 2016. Secondly, they argued that there is no proof of the causal link between the infringement and the price increase. The appellate court referred to the Court of Justice a number of questions on the scope ratione temporis of certain provisions of the Damages Directive concerning the limitation period, the assessment of harm, as well as the compatibility of the national legislation applicable to actions for damages, in view of Article 101 TFEU and the principle of effectiveness (effet utile).

AG Rantos’s Opinion, rendered on October 28, methodically treated these questions one by one.

The main question is a comparative lawyer’s paradise. Are the specific rules of the Damages Directive at stake “procedural” or “substantive” in nature? This is an important question because, under Article 22 of the Damages Directive, procedural rules apply to actions that have been brought after the Directive entered into force (on December 26, 2014), but substantive provisions do not apply retroactively to “situations existing” before that date. And the Advocate General rightly observes that Article 22 does not clarify what constitutes “substance” and what is “procedure” (para. 43). To make things worse, the Court’s rulings in Cogeco and Skanska do not give an answer to the question whether the reference to “situations existing” means (i) the infringement itself, (ii) the decision of the competition authority that finds the infringement or (iii) the filing of an action for damages (para. 39 of the Opinion). So this is quite fertile territory for the Advocate General.

The Opinion starts from the premise that, in order to ensure a consistent and uniform application of EU competition law, determining which provisions of the Directive are substantive or procedural must be assessed in the light of EU law and not in the light of national law (paras 58-59). And then come the concrete proposals of AG Rantos.

First, on the nature of the rules on limitation periods (Article 10 of the Directive), the Advocate General concludes that these are “substantive”, because such periods have the function of ensuring the protection both of the injured party, who must have sufficient time in which to gather the appropriate information with a view to a possible action, and the person liable for the damage, by preventing the injured party from being able to delay indefinitely the exercise of his or her right to damages (para. 66). Thus, the five-year limitation period of the Damages Directive does not apply here, although the action was brought after the Directive and the national transposing measures entered into force (May 26, 2017). The critical element is that the action concerns facts occurring and penalties imposed before that date.

Second, on the nature of the presumption that cartels cause harm (Article 17(2) of the Directive), the Advocate General also considers the relevant provision as “substantive”. According to him, “by allocating the burden of proof to the infringer and thus exempting the injured party from the obligation of proving the existence of the harm suffered on account of the cartel or the causal link between the cartel and the harm, that presumption is directly linked to the allocation of non-contractual civil liability to the infringer concerned and, as a result, directly affects the latter’s legal situation” (para. 81 – unofficial translation). Thus, this rule cannot apply to infringements committed before the Directive and the national transposing measures entered into force.

Third, by contrast, Article 17(1) of the Directive and the national transposing measures entered into force concerning the courts’ power to assess the harm are “procedural” and can apply retroactively, i.e. to harm suffered as a result of a competition law infringement that ceased before the entry into force of the national transposing legislation, if the action for damages was brought after that date (para. 77).

Having answered the above questions in that way, the Advocate General examines, next, whether the rules governing non-contractual liability laid down in the Spanish Civil Code are compatible with the EU principle of effectiveness, according to which any person who has suffered harm must be able to seek compensation for loss. He starts from the premise that “the principle of effectiveness cannot impose the retroactive application of the substantive provisions of Directive 2014/104. This would contradict general principles of law, such as the principle of legal certainty” (para. 91 – unofficial translation). As regards the duration of the limitation period, the Advocate General points out, while acknowledging that the one-year period provided for in the Spanish legislation is significantly shorter than the five-year period provided for in the Directive, that other elements of the national rules on limitation must also be taken into consideration (para. 101). For the rest, the Advocate General refers to Cogeco. So the one-year duration is not problematic per se.

As regards the dies a quo for calculating the one-year limitation period laid down in the Civil Code, the Advocate General considers that that period begins to run from the date of publication of the summary of the Commission decision in the Official Journal of the EU, in this case, April 6, 2017. This means that the action for damages brought by the claimant on April 1, 2018 is not time-barred. The Advocate General proposes the rejection of the defendants’ argument that the period may begin to run from the date of publication of the Commission press release. Indeed, the mere publication of that document does not enable the injured party concerned to know all the information required to exercise his or her right to bring an action for damages. In addition, the Advocate General points out that “there is no general duty of due diligence for victims of competition law infringements requiring them to monitor the publication of such press releases” (para. 118 – unofficial translation). Interestingly, in para. 124, the Advocate General criticizes the Commission’s practice not to publish a summary in the Official Journal at the time of the formal adoption of the infringement decision. In his view, it is rather easy to prepare a very short non-confidential test for publication in the Official Journal, echoing the information that is already provided with the press release, anyway.

Lastly, the Advocate General emphasizes that the fact that the presumption of harm provided for in the Directive does not to apply to the present case does not preclude national courts from applying similar presumptions relating to the burden of proof contained in national law, as long as such presumptions and rules do not offend against the EU general principles of effectiveness and equivalence (paras 140-141).

In short, a very thorough analysis – almost of pedagogical value.

  1. AG Bobek’s non bis in idem Opinions

On September 2, 2021, Advocate General Bobek delivered two very important Opinions on the application of the principle of non bis in idem. The first Opinion, in bpost, relates to the relationship between competition law and regulation, and the second Opinion, in Nordzucker, has to do with the relationship between national and EU competition enforcement. The Opinions were eagerly awaited because of their relevance to the future relationship between the DMA and competition law enforcement (see here).

As the Advocate General himself proposes, the reader should first start from the bpost Opinion, which includes a more thorough analysis of the principle of non bis in idem and his main proposal on how to interpret this principle under EU law. Indeed, the Nordzucker Opinion relies heavily on that analysis.

So we start with bpost first. The case was about whether the principle of non bis in idem applies between regulatory and competition proceedings, where the same facts may be examined under two different legal regimes. In that case, the Belgian Competition Authority (BCA) imposed a fine on bpost, the historical postal service provider in Belgium, for abuse of dominance. According to the BCA, the abuse related to quantity discounts offered by bpost that encouraged major clients to contract directly with it, thus placing consolidators (who supply senders with routing services upstream of the postal distribution service) at a competitive disadvantage. The complication here was that a year earlier the Belgian regulatory authority for postal services (IBPT) had also found an infringement of the regulatory regime based on the same or similar facts and had imposed a fine. In the end, that finding was annulled on appeal, but the fact remained that there were two consecutive proceedings leading to fines for the same facts.

After a long process of appeals and remands, the finding of the competition infringement came back to the Brussels Court of Appeal, which referred the non bis in idem question to the Court of Justice. The Belgian court asked whether the principle of non bis in idem applies to this situation, although the case relates to two infringements of different legal regimes: regulation and competition law.

The fact that these cases are heard by the Grand Chamber may mean that the Court is willing to re-examine its Toshiba line of case law, which follows a rather restrictive interpretation of double jeopardy in competition cases, as opposed to other areas of EU law. Indeed, in areas other than competition law, the Court of Justice assesses “idem”, i.e. the identity of an offence, on the basis of only a two-fold criterion, identity of facts and offender (see case law cited by AG Wahl’s Opinion in Powszechny Zakład, para. 25). In competition cases, however, the Court of Justice employs an additional criterion, identity of the legal interest protected (Aalborg Portland, para. 338; Toshiba, para. 97, although in that case AG Kokott had urged the Court to align the position in EU law and dispense with the third condition, without success).

The bpost Opinion explains all this in detail. The readers that do not possess infinite time may wish to start reading from para. 126 onwards. This is where most of the substance is. AG Bobek proposes a repositioning of the case law, which has given rise to a “fragmented and partially contradictory mosaic of parallel regimes” (para. 6), but not in the direction one would have expected. Rather than aligning the competition case law with the case law in other areas of EU law, he proposes the reverse! In other words, the Advocate General proposes “making the examination of the protected legal interests, and thus of the objective pursued, part of the consideration of idem” (para. 132). For him, “the assessment of idem for the purposes of Article 50 of the Charter should rely on a triple identity: of the offender, of the relevant facts, and of the protected legal interest” (para. 133). The Advocate General reaches that conclusion after a critical analysis of the case law, which shows, in his view, that “even where the idem factum approach has been adhered to, the concept of protected legal interest has in fact never really disappeared. Much like a little chameleon, it simply adopted different colours, gluing itself to the different sticks or branches that were available within each line of case-law at the given time” (para. 128).

So, if that is the test, does the principle of non bis in idem apply to the facts of the case? AG Bobek conducts an ad hoc analysis and finds that this is not the case. In his words, “subject to verification by the referring court, both offences that have been pursued successively in the sectoral and competition proceedings seem to be linked to the protection of a different legal interest and to a legislation pursuing a different objective. First, in terms of the protected legal interest, achieving liberalisation of certain, previously monopolistic, markets follows a different logic than the ongoing and horizontal protection of competition. Second, that is also evident with regard to the undesirable consequences that punishment of each of the offences is intended to prevent. If the aim is to liberalise a sector, then potential harm caused to competition upstream or downstream is not necessarily an issue that the sectoral regulatory framework must tackle. By contrast, an abuse of a dominant position that results in a distortion of competition upstream or downstream from the dominant undertaking is very much a concern of competition rules” (para. 162).

This is a very interesting line of reasoning. If the Court of Justice were to follow it, it would mean that non bis in idem would not stand in the way of cumulative proceedings under competition law and sectoral regulation in the energy, post and telecoms sectors, but I do not think the rationale is transposable to the DMA. The latter is certainly not about “achieving liberalisation of certain, previously monopolistic, markets” and I am not at all sure that it “follows a different logic than the ongoing and horizontal protection of competition. Besides, the substantive provisions in Articles 5 and 6 of the DMA are precisely inspired by existing competition law investigations and decisions. In addition, the two guiding principles behind the DMA, fairness and contestability, are utterly linked to competition law. In my own view, the DMA is nothing more than sui generis competition law. The ex ante parameter is not critical and, indeed, AG Bobek does not refer to it at all.

Let us now go to the Nordzucker Opinion. Here, the Advocate General’s Opinion goes to the opposite direction in terms of the final proposal, i.e. he recommends the Court to find that the principle of non bis in idem applies to the facts of the case. This was a preliminary reference from Austria and relates to a cartel fine imposed by the Austrian competition authority on an undertaking in the sugar sector. However, that undertaking had already been fined by the Federal Cartel Office in Germany exactly for the same facts. A further complication of this case is that there is a disagreement as to whether the German decision had taken into account or not the effects of the infringement in Austria.

As already explained, the Nordzucker Opinion builds on bpost. The most interesting part of the Opinion starts in para. 44: “Do EU and national competition laws protect the same legal interest?” The Advocate General, in effect, answers “yes”, albeit with a few nuances.

The Opinion’s starting point is that there is no doubt that national and EU competition law have massively converged since 1969, when the Court of Justice had rendered its Walt Wilhelm ruling (paras 45, 47). At the same time, the Advocate General – in my view very rightly – dismisses the “geographical scope” as a critical distinctive feature between EU and national law. Rather, it is the “normative quality of the interest (or objective) pursued” that should count (para. 52). Put eloquently:

In other words, I do not believe that the mere (quantitative) difference in the territorial scope of the same infringement, and thus of the given rule, is per se indicative of a (qualitative) difference in the legal interest. While EU competition law covers situations in which trade between the Member States is affected, national competition law applies to internal situations. In my view, that difference points to the territorial extent of the infringement, possibly coupled with the gravity of the interference with the protected legal interest, but not necessarily to the different quality of that protected legal interest” (para. 53).

However, the answer to the main question at stake cannot be given in the abstract, but as the Opinion rightly says, “[t]he issue of the protected legal interest ought to be assessed with regard to a specific provision. It must focus on the specific interest or purpose that the provision being applied pursues, what that provision penalises and why” (para. 44). Then, the Advocate General interestingly relies on Article 3 of Regulation 1/2003 and builds a rule based on three different scenarios (para. 51): (i) application of Article 101 TFEU and of the equivalent national provision (Article 3(2), first sentence); (ii) application of Article 102 TFEU and of the equivalent national provision on abuse of dominance (Article 3(2), second sentence); and (iii) application of EU competition law and of national rules pursuing other objectives than Articles 101 and 102 TFEU (Article 3(3)).

At this point, the Advocate General’s analysis becomes rather elliptical. It refers in reality only to scenario (i) and clarifies that in that case we are dealing with the “same legal interest”, so the non bis in idem principle would apply. It is not clear what the Advocate General thinks of the other two scenarios. I think it is implicit that in scenario (iii), we would not be dealing with the same legal interest, but there is complete lack of clarity as to scenario (ii), although the Advocate General does say that “there is a rather large but not complete substantive overlap for situations falling under Article 102 TFEU, in which the Member States may adopt stricter rules” (para. 51). So I think the Advocate General says that in the case of application of unilateral conduct rules, it will be very much an ad hoc issue. Possibly, if the national rule is a copy-paste of Article 102 TFEU, then the principle of non bis in idem would apply. But if it is quite different, then it may not apply.

A separate legal question that arose in this case concerns the principle of territoriality. In the case at hand, it was unclear whether the German Federal Cartel Office had taken into account the effects of the conduct in question on the Austrian market. This was a factual issue and the Advocate General defers to the national court. If that was the case, then clearly such a fact would be relevant for the examination of the applicability of the principle of non bis in idem (para. 87). However, is such “extraterritoriality” possible in the first place? Here, the Advocate General has some interesting things to say. He rejects the view that such an “extraterritorial” power is given by Regulation 1/2003 (para. 78 et seq.). So, in his view, “for the powers of an NCA to be exercised extraterritorially, there must be an adequate legal basis which, as EU law currently stands, can only come from the national legal system” (para. 85). In this sense, the Advocate General disagrees with views long-held by Wouter Wils, who has argued that such a power – or indeed duty – for NCAs is inherent in Regulation 1/2003 and the principle of effectiveness of EU competition law. We will see if the Court goes into this particular issue – I have my doubts.

These were among the last Opinions of AG Bobek, who will be missed, since his mandate expired in September. One thing is for sure: the forthcoming judgments will come from the Court’s Grand Chamber, so we should expect important pronouncements that will clarify the state of affairs.

II. General Court

  1. Amazon

On October 14, the General Court rejected as inadmissible a challenge by Amazon of the Commission’s opening of proceedings in the BuyBox case. This was a rather peculiar case. The facts of the case are as follows. On November 10, 2020, the Commission opened an antitrust investigation into Amazon’s business practices that, in the Commission’s view, might artificially favour its own retail offers and offers of marketplace sellers that use Amazon’s logistics and delivery services. However, in the decision opening proceedings, which was based on Article 11(6) of Regulation 1/2003, the Commission carved the territory of Italy out and opened proceedings for the whole of the EEA excluding Italy. According to the Commission’s press release, the carve-out was due to the fact that “the Italian Competition Authority started to investigate partially similar concerns last year, with a particular focus on the Italian market”. So the carve-out in effect preserved the competence of the Italian authority to continue its investigation and Amazon is subject to two parallel investigations, one by the Commission and another by the Italian Competition Authority. Amazon challenged the Commission’s opening of proceedings, to the extent Italy was carved-out and requested the General Court to annul partially the Commission’s decision to the above extent. This was because, in Amazon’s view, the Commission violated the letter and spirit of Article 11(6), which does not allow for such carve-outs.

The Commission, however, submitted a plea of inadmissibility, which the General Court upheld, thus rejecting by Order Amazon’s application for annulment as inadmissible. If you expect me to offer some critical comment on this extremely interesting case, you will be disappointed. I was representing Amazon, so my lips are sealed 🙂 – so read it yourselves and you can send me your own views by email… I will only give below the General Court’s reasoning.

The General Court held that the exclusion of Italy from the geographical scope of the opening of proceedings does not alter Amazon’s rights and obligations in substantive terms, nor does it adversely affect its procedural rights (para. 31). The General Court did accept that, under the Slovak Telekom case law, Article 11(6) of Regulation 1/2003 provides the undertakings with a protection against parallel proceedings, but held that such a protection does not imply any right, for the benefit of an undertaking, to have a case dealt with in its entirety by the Commission. So, “the fact that Amazon must defend itself before two different authorities does not have any effects other than procedural and thus does not affect its legal situation” (para. 36). Therefore, the General Court upheld the Commission’s objection of inadmissibility and held that the exclusion of Italy from the territorial scope of the investigation constituted a mere preparatory act which does not produce legal effects vis-à-vis Amazon within the meaning of Article 263 TFEU.

  1. Altice

This was an appeal against a Commission decision imposing two fines on Altice for (i) failure to notify its acquisition of Portugal Telecom prior to implementation (Article 4 of the Merger Regulation) and (ii) gun-jumping, i.e. implementing a transaction prior to the clearance decision (Article 7 of the Merger Regulation). The General Court confirmed the Commission’s decision in its substance but slightly reduced the fine imposed for the failure to notify, because Altice had informed the Commission of the transaction before the SPA was signed and, immediately after signing, it had sent to the Commission a case team allocation request.

The judgment is quite long and deserves a closer look, so I limit myself only to a few main points.

First, Altice had argued that the gun-jumping violation and fine rendered redundant the failure to notify violation and fine. This argument was rejected by the General Court on the basis that Articles 4(1) and 7(1) of the Merger Regulation pursue autonomous objectives (paras 56, 60). Article 4(1) requires undertakings to notify a concentration before it is implemented, while Article 7(1) prevents undertakings from implementing a concentration before the Commission has cleared it. In addition, Article 4(1) lays down a positive obligation to act, whereas Article 7(1) lays down a negative obligation not to act. Furthermore, an infringement of Article 4(1) is an instantaneous infringement, whereas an infringement of Article 7(1) is a continuous infringement (para. 58).

Second, the General Court rejected Altice’s arguments about the breach of the principles of proportionality and non bis in idem, because of the dual fine. There was nothing unlawful in the imposition of two fines, which was consistent with the objective of the Merger Regulation to ensure effective control of concentrations. Besides, accepting Altice’s arguments would mean that the Commission would not be able to distinguish between a situation in which the undertaking complies with the notification obligation but infringes the standstill obligation, and a situation in which the undertaking infringes both obligations (para. 63).

In terms of substance, the General Court agreed with the Commission that Altice had exercised decisive influence on Portugal Telecom and had implemented the merger before it was cleared. And, in any event, if it were true that Altice’s intervention in Portugal Telecom’s affairs were justified on account of the unusual nature of the operations concerned, it should have requested a derogation from the standstill obligation, on the basis of Article 7(3) of the Merger Regulation, which Altice never did (paras 216-217).

  1. Capacitors

A typical series of appeals against a Commission decision in the capacitors cartel case gave rise to some interesting judgments. Of the five judgments, the most interesting ones are Nichicon and Nippon Chemi-Con. The applicants contested the Commission’s jurisdiction on the ground that the cartel was Asia-oriented and was not implemented in the EEA. The General Court recalled that the EU has jurisdiction in two situations: (i) when the agreements in question are implemented in the territory of the internal market, irrespective of the place where they were concluded; and (ii) when it is foreseeable that those practices will have an immediate and substantial effect in the internal market. The General Court concluded that the criterion of the implementation of the cartel was satisfied in the present case and the Commission had jurisdiction.

In Nichicon, the applicant also raised a non bis in idem argument (it seems this is currently a hot topic in Luxembourg 🙂). It argued that the Commission had infringed that principle, together with the principle of proportionality, because other non-EU authorities had also imposed cartel fines. The General Court, however, based on long-standing case law, rejected this argument, because the principle non bis in idem cannot apply as between the Commission and non-Member State authorities, since the respective competition laws do not pursue the same objectives. The condition of unity of the legal interest protected, which is necessary for the application of the principle non bis in idem, was therefore lacking. As regards the principle of proportionality, the General Court held that the Commission has discretion as to whether to take fines imposed by non-EU authorities into account. Accordingly, although it cannot be ruled out that the Commission may take into account fines imposed previously by the authorities of non-Member States, it cannot be required to do so.

III. European Commission

On the Commission front, I consider noteworthy two decisions.

  1. Car emissions

This was a widely anticipated decision. In July, the Commission issued its infringement decision against BMW and the Volkswagen group (Volkswagen, Audi and Porsche), imposing fines totalling €875m, for participating in a cartel aimed at restricting competition regarding the development of technology to clean the emissions of diesel passenger cars.

According to the Commission, the companies held regular technical meetings to discuss the development of the selective catalytic reduction (SCR)-technology which eliminates harmful nitrogen oxide (NOx)-emissions from diesel passenger cars through the injection of urea (also called “AdBlue”) into the exhaust gas stream. During these meetings, and for over five years, the car manufacturers colluded to avoid competition on cleaning better than what is required by law despite the relevant technology being available. They also exchanged commercially sensitive information on these questions.

The three companies acknowledged their involvement. Daimler received full immunity for revealing the existence of the cartel, while BMW and Volkswagen received reduced fines in return for cooperating with the Commission’s investigation and for agreeing to settle the case.

  1. Illumina/Grail

As a reminder, this is the first case where the Commission intervened after changing its “approach” with the publication of its new Guidance on Article 22 of the Merger Regulation. This is about mergers that fall below EU and national notification thresholds and therefore there is neither EU nor national jurisdiction in the first place. Whereas until now, the Commission was discouraging Article 22 referrals of such transactions from Member States, from now on, it will welcome such referrals in appropriate cases. The Illumina/Grail transaction was the first “victim”. On April 19, the Commission “accepted” pursuant to Article 22(3) of the Merger Regulation a referral under Article 22(1) made by the French Competition Authority, and asserted its jurisdiction to examine the concentration. On July 22, the Commission opened an in-depth investigation into the effects of the transaction, but shortly afterwards, on August 18, while the Commission’s review was ongoing, Illumina publicly announced that it had completed its acquisition.

To say that this was something that the Commission (and Executive Vice-President Vestager) did not like, would be an understatement🙂. So the whole Commission mechanism was unleashed. The Commission decided to punish this behaviour and, on October 29, adopted an interim measures decision. According to the Commission, the interim measures aim to prevent the potentially irreparable detrimental impact of the transaction on competition as well as the possible irreversible integration of the merging parties, pending the outcome of the Commission’s merger investigation. In particular, the interim measures adopted provide that:

  • Grail shall be kept separate from Illumina and be run by (an) independent Hold Separate Manager(s), exclusively in the interest of Grail (and not of Illumina).
  • Illumina and Grail are prohibited from sharing confidential business information, except where the disclosure is required to comply with the law or in line with the ordinary course of their supplier-customer relationship.
  • Illumina has the obligation to finance additional funds necessary for the operation and development of Grail.
  • The business interactions between the parties shall be undertaken at arm’s length, in line with industry practice, hence without unduly favouring Grail to the detriment of its competitors.
  • Grail shall actively work on alternative options to the transaction to prepare for the possible scenario in which the deal would have to be undone in case the Commission were to declare the transaction incompatible with the internal market. [This is a very interesting interim measure and I wonder how it can be implemented…]

In parallel, the Commission will continue (i) its investigation in whether the implementation of the transaction amounts to gun-jumping and (ii) its in-depth merger control investigation.

And at some point the General Court will have to tell us whether the EU had jurisdiction in the first place. The moment of truth will not take long, as the Court has granted the expedited procedure to this matter…


Citation: Makis Komninos, Competition Stories: July to October 2021, CONCURRENTIALISTE (November 9, 2021)

Read all the competition stories over here: link

Related Posts